[Freedreno] [PATCH 12/14] drm/msm/dpu: remove topology name

Jeykumar Sankaran jsanka at codeaurora.org
Tue Sep 4 23:03:25 UTC 2018


On 2018-08-31 09:08, Sean Paul wrote:
> On Tue, Aug 28, 2018 at 05:40:01PM -0700, Jeykumar Sankaran wrote:
>> Strip down the support for topology enums. It
>> can be replaced with simple hw count checks.
>> 
> 
> Changes in v4:
> ...
> 
>> Signed-off-by: Jeykumar Sankaran <jsanka at codeaurora.org>
>> ---
>>  drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c        |  3 --
>>  drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.h        |  1 -
>>  drivers/gpu/drm/msm/disp/dpu1/dpu_encoder_phys.h   |  9 ++++--
>>  .../gpu/drm/msm/disp/dpu1/dpu_encoder_phys_vid.c   |  7 +++--
>>  drivers/gpu/drm/msm/disp/dpu1/dpu_rm.c             | 36
> ++++++----------------
>>  drivers/gpu/drm/msm/disp/dpu1/dpu_rm.h             | 24 
>> ---------------
>>  6 files changed, 19 insertions(+), 61 deletions(-)
>> 
>> diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c
> b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c
>> index dbf669e..56ef349 100644
>> --- a/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c
>> +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c
>> @@ -1013,7 +1013,6 @@ static void dpu_encoder_virt_mode_set(struct
> drm_encoder *drm_enc,
>>  	struct drm_connector *conn = NULL, *conn_iter;
>>  	struct dpu_rm_hw_iter pp_iter, ctl_iter;
>>  	struct msm_display_topology topology;
>> -	enum dpu_rm_topology_name topology_name;
>>  	struct dpu_hw_ctl *hw_ctl[MAX_CHANNELS_PER_ENC] = { NULL };
>>  	int i = 0, ret;
>> 
>> @@ -1069,7 +1068,6 @@ static void dpu_encoder_virt_mode_set(struct
> drm_encoder *drm_enc,
>>  		hw_ctl[i] = (struct dpu_hw_ctl *)ctl_iter.hw;
>>  	}
>> 
>> -	topology_name = dpu_rm_get_topology_name(topology);
>>  	for (i = 0; i < dpu_enc->num_phys_encs; i++) {
>>  		struct dpu_encoder_phys *phys = dpu_enc->phys_encs[i];
>> 
>> @@ -1089,7 +1087,6 @@ static void dpu_encoder_virt_mode_set(struct
> drm_encoder *drm_enc,
>>  			phys->hw_ctl = hw_ctl[i];
>> 
>>  			phys->connector = conn->state->connector;
>> -			phys->topology_name = topology_name;
>>  			if (phys->ops.mode_set)
>>  				phys->ops.mode_set(phys, mode, adj_mode);
>>  		}
>> diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.h
> b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.h
>> index 60f809f..c5600e6 100644
>> --- a/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.h
>> +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.h
>> @@ -35,7 +35,6 @@
>>   * @needs_cdm:	Encoder requests a CDM based on pixel format
> conversion needs
>>   * @display_num_of_h_tiles: Number of horizontal tiles in case of 
>> split
>>   *                          interface
>> - * @topology:   Topology of the display
>>   */
>>  struct dpu_encoder_hw_resources {
>>  	enum dpu_intf_mode intfs[INTF_MAX];
>> diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder_phys.h
> b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder_phys.h
>> index b3917e0..b5fc65c 100644
>> --- a/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder_phys.h
>> +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder_phys.h
>> @@ -24,6 +24,7 @@
>>  #include "dpu_hw_top.h"
>>  #include "dpu_hw_cdm.h"
>>  #include "dpu_encoder.h"
>> +#include "dpu_crtc.h"
>> 
>>  #define DPU_ENCODER_NAME_MAX	16
>> 
>> @@ -213,7 +214,6 @@ struct dpu_encoder_irq {
>>   * @split_role:		Role to play in a split-panel
> configuration
>>   * @intf_mode:		Interface mode
>>   * @intf_idx:		Interface index on dpu hardware
>> - * @topology_name:	topology selected for the display
>>   * @enc_spinlock:	Virtual-Encoder-Wide Spin Lock for IRQ purposes
>>   * @enable_state:	Enable state tracking
>>   * @vblank_refcount:	Reference count of vblank request
>> @@ -243,7 +243,6 @@ struct dpu_encoder_phys {
>>  	enum dpu_enc_split_role split_role;
>>  	enum dpu_intf_mode intf_mode;
>>  	enum dpu_intf intf_idx;
>> -	enum dpu_rm_topology_name topology_name;
>>  	spinlock_t *enc_spinlock;
>>  	enum dpu_enc_enable_state enable_state;
>>  	atomic_t vblank_refcount;
>> @@ -361,11 +360,15 @@ struct dpu_encoder_phys
> *dpu_encoder_phys_cmd_init(
>>  static inline enum dpu_3d_blend_mode
> dpu_encoder_helper_get_3d_blend_mode(
>>  		struct dpu_encoder_phys *phys_enc)
>>  {
>> +	struct dpu_crtc_state *dpu_cstate;
>> +
>>  	if (!phys_enc || phys_enc->enable_state == DPU_ENC_DISABLING)
>>  		return BLEND_3D_NONE;
>> 
>> +	dpu_cstate = to_dpu_crtc_state(phys_enc->parent->crtc->state);
>> +
>>  	if (phys_enc->split_role == ENC_ROLE_SOLO &&
>> -	    phys_enc->topology_name == DPU_RM_TOPOLOGY_DUALPIPE_3DMERGE)
>> +	    (dpu_cstate->num_mixers == 2))
> 
> I guess this should be CRTC_DUAL_MIXERS? Perhaps you should add a 
> helper
> called
> dpu_crtc_state_is_stereo() instead of littering these checks 
> everywhere.
> 
>>  		return BLEND_3D_H_ROW_INT;
>> 
>>  	return BLEND_3D_NONE;
>> diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder_phys_vid.c
> b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder_phys_vid.c
>> index 6de13f4..1643e2f 100644
>> --- a/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder_phys_vid.c
>> +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder_phys_vid.c
>> @@ -355,13 +355,14 @@ static void 
>> dpu_encoder_phys_vid_underrun_irq(void
> *arg, int irq_idx)
>> 
>>  static bool _dpu_encoder_phys_is_dual_ctl(struct dpu_encoder_phys
> *phys_enc)
>>  {
>> +	struct dpu_crtc_state *dpu_cstate;
>> +
>>  	if (!phys_enc)
>>  		return false;
>> 
>> -	if (phys_enc->topology_name == DPU_RM_TOPOLOGY_DUALPIPE)
>> -		return true;
>> +	dpu_cstate = to_dpu_crtc_state(phys_enc->parent->crtc->state);
>> 
>> -	return false;
>> +	return (dpu_cstate->num_ctls > 1);
> 
> Unnecessary parens
> 
>>  }
>> 
>>  static bool dpu_encoder_phys_vid_needs_single_flush(
>> diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_rm.c
> b/drivers/gpu/drm/msm/disp/dpu1/dpu_rm.c
>> index 13c0a36..fbd489f 100644
>> --- a/drivers/gpu/drm/msm/disp/dpu1/dpu_rm.c
>> +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_rm.c
>> @@ -33,7 +33,6 @@
>>  				(t).num_intf == (r).num_intf)
>> 
>>  struct dpu_rm_topology_def {
>> -	enum dpu_rm_topology_name top_name;
>>  	int num_lm;
>>  	int num_comp_enc;
>>  	int num_intf;
>> @@ -42,10 +41,10 @@ struct dpu_rm_topology_def {
>>  };
>> 
>>  static const struct dpu_rm_topology_def g_top_table[] = {
>> -	{   DPU_RM_TOPOLOGY_NONE,                 0, 0, 0, 0, false },
>> -	{   DPU_RM_TOPOLOGY_SINGLEPIPE,           1, 0, 1, 1, false },
>> -	{   DPU_RM_TOPOLOGY_DUALPIPE,             2, 0, 2, 2, true  },
>> -	{   DPU_RM_TOPOLOGY_DUALPIPE_3DMERGE,     2, 0, 1, 1, false },
>> +	{   0, 0, 0, 0, false },
>> +	{   1, 0, 1, 1, false },
>> +	{   2, 0, 2, 2, true  },
>> +	{   2, 0, 1, 1, false },
> 
> You don't need this. The data it provides can be deduced already:
> num_ctl == num_intf
> needs_split_display == (num_intf > 1)
> 
>>  };
>> 
>>  /**
>> @@ -72,13 +71,11 @@ struct dpu_rm_requirements {
>>   * @enc_id:	Reservations are tracked by Encoder DRM object ID.
>>   *		CRTCs may be connected to multiple Encoders.
>>   *		An encoder or connector id identifies the display path.
>> - * @topology	DRM<->HW topology use case
>>   */
>>  struct dpu_rm_rsvp {
>>  	struct list_head list;
>>  	uint32_t seq;
>>  	uint32_t enc_id;
>> -	enum dpu_rm_topology_name topology;
>>  };
>> 
>>  /**
>> @@ -122,8 +119,8 @@ static void _dpu_rm_print_rsvps(
>>  	DPU_DEBUG("%d\n", stage);
>> 
>>  	list_for_each_entry(rsvp, &rm->rsvps, list) {
>> -		DRM_DEBUG_KMS("%d rsvp[s%ue%u] topology %d\n", stage,
> rsvp->seq,
>> -			      rsvp->enc_id, rsvp->topology);
>> +		DRM_DEBUG_KMS("%d rsvp[s%ue%u]\n", stage, rsvp->seq,
>> +			      rsvp->enc_id);
>>  	}
>> 
>>  	for (type = 0; type < DPU_HW_BLK_MAX; type++) {
>> @@ -146,18 +143,6 @@ struct dpu_hw_mdp *dpu_rm_get_mdp(struct dpu_rm
> *rm)
>>  	return rm->hw_mdp;
>>  }
>> 
>> -enum dpu_rm_topology_name
>> -dpu_rm_get_topology_name(struct msm_display_topology topology)
>> -{
>> -	int i;
>> -
>> -	for (i = 0; i < DPU_RM_TOPOLOGY_MAX; i++)
>> -		if (RM_IS_TOPOLOGY_MATCH(g_top_table[i], topology))
>> -			return g_top_table[i].top_name;
>> -
>> -	return DPU_RM_TOPOLOGY_NONE;
>> -}
>> -
>>  void dpu_rm_init_hw_iter(
>>  		struct dpu_rm_hw_iter *iter,
>>  		uint32_t enc_id,
>> @@ -760,7 +745,6 @@ static int _dpu_rm_make_next_rsvp(
>>  	/* Create reservation info, tag reserved blocks with it as we go
> */
>>  	rsvp->seq = ++rm->rsvp_next_seq;
>>  	rsvp->enc_id = enc->base.id;
>> -	rsvp->topology = reqs->topology->top_name;
>>  	list_add_tail(&rsvp->list, &rm->rsvps);
>> 
>>  	ret = _dpu_rm_reserve_lms(rm, rsvp, reqs);
>> @@ -807,7 +791,7 @@ static int _dpu_rm_populate_requirements(
>> 
>>  	dpu_encoder_get_hw_resources(enc, &reqs->hw_res, conn_state);
>> 
>> -	for (i = 0; i < DPU_RM_TOPOLOGY_MAX; i++) {
>> +	for (i = 0; i < ARRAY_SIZE(g_top_table); i++) {
>>  		if (RM_IS_TOPOLOGY_MATCH(g_top_table[i],
>>  					req_topology)) {
>>  			reqs->topology = &g_top_table[i];
> 
> Since you can get rid of g_top_table, you can get rid of reqs->topology
> too.
> 
DPU RM is using reqs->topology (ie. struct dpu_rm_topology_def) to 
represent
the hw_block requirements for a display. Removing the structure
might require changes in rest of RM alloc functions.

If you can hold off on this, I am planning to address your comment in 
the
RM refactor series.

Jeykumar S.

>> @@ -831,9 +815,8 @@ static int _dpu_rm_populate_requirements(
>> 
>>  	DRM_DEBUG_KMS("top_ctrl: 0x%llX num_h_tiles: %d\n",
> reqs->top_ctrl,
>>  		      reqs->hw_res.display_num_of_h_tiles);
>> -	DRM_DEBUG_KMS("num_lm: %d num_ctl: %d topology: %d split_display:
> %d\n",
>> +	DRM_DEBUG_KMS("num_lm: %d num_ctl: %d split_display: %d\n",
>>  		      reqs->topology->num_lm, reqs->topology->num_ctl,
>> -		      reqs->topology->top_name,
>>  		      reqs->topology->needs_split_display);
>> 
>>  	return 0;
>> @@ -969,8 +952,7 @@ static int _dpu_rm_commit_rsvp(
>>  	}
>> 
>>  	if (!ret)
>> -		DRM_DEBUG_KMS("rsrv enc %d topology %d\n", rsvp->enc_id,
>> -			      rsvp->topology);
>> +		DRM_DEBUG_KMS("rsrv enc %d\n", rsvp->enc_id);
> 
> This debug message doesn't provide any value any longer.
> 
>> 
>>  	return ret;
>>  }
>> diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_rm.h
> b/drivers/gpu/drm/msm/disp/dpu1/dpu_rm.h
>> index ffd1841..6fd9e7b 100644
>> --- a/drivers/gpu/drm/msm/disp/dpu1/dpu_rm.h
>> +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_rm.h
>> @@ -21,21 +21,6 @@
>>  #include "dpu_hw_top.h"
>> 
>>  /**
>> - * enum dpu_rm_topology_name - HW resource use case in use by 
>> connector
>> - * @DPU_RM_TOPOLOGY_NONE:                 No topology in use 
>> currently
>> - * @DPU_RM_TOPOLOGY_SINGLEPIPE:           1 LM, 1 PP, 1 INTF/WB
>> - * @DPU_RM_TOPOLOGY_DUALPIPE:             2 LM, 2 PP, 2 INTF/WB
>> - * @DPU_RM_TOPOLOGY_DUALPIPE_3DMERGE:     2 LM, 2 PP, 3DMux, 1 
>> INTF/WB
>> - */
>> -enum dpu_rm_topology_name {
>> -	DPU_RM_TOPOLOGY_NONE = 0,
>> -	DPU_RM_TOPOLOGY_SINGLEPIPE,
>> -	DPU_RM_TOPOLOGY_DUALPIPE,
>> -	DPU_RM_TOPOLOGY_DUALPIPE_3DMERGE,
>> -	DPU_RM_TOPOLOGY_MAX,
>> -};
>> -
>> -/**
>>   * enum dpu_rm_topology_control - HW resource use case in use by
> connector
>>   * @DPU_RM_TOPCTL_RESERVE_LOCK: If set, in AtomicTest phase, after a
> successful
>>   *                              test, reserve the resources for this
> display.
>> @@ -187,13 +172,4 @@ void dpu_rm_init_hw_iter(
>>   */
>>  int dpu_rm_check_property_topctl(uint64_t val);
>> 
>> -/**
>> - * dpu_rm_get_topology_name - returns the name of the the given
> topology
>> - *                            definition
>> - * @topology: topology definition
>> - * @Return: name of the topology
>> - */
>> -enum dpu_rm_topology_name
>> -dpu_rm_get_topology_name(struct msm_display_topology topology);
>> -
>>  #endif /* __DPU_RM_H__ */
>> --
>> The Qualcomm Innovation Center, Inc. is a member of the Code Aurora
> Forum,
>> a Linux Foundation Collaborative Project
>> 

-- 
Jeykumar S


More information about the Freedreno mailing list